Date:         Mon, 21 Mar 2011 22:41:14 -0400
Reply-To:     R B <ryan.andrew.black@gmail.com>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         R B <ryan.andrew.black@gmail.com>
Subject:      Re: Estimating Event History Model using SPSS or STATA
Comments: To: =?UTF-8?B?5byg5Lym?= <zhanglun1525@gmail.com>
In-Reply-To:  <AANLkTi=7hr0DPPJKrVabBSWV4aOGaH4phhzJgXngbw=g@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8

There is a generalized linear mixed model procedure (GENLINMIXED) in
SPSS 19. However, I doubt an event history model can be fit using this
procedure. You might want to investigate the very flexible nonlinear
mixed model procedure (NLMIXED) in SAS to fit a multilevel event
history model. I'm fairly certain the NLMIXED procedure can handle
such a model, though writing the appropriate code could be tricky.

Ryan

2011/3/21 å¼ ä¼¦ <zhanglun1525@gmail.com>:
> Dear all:
> IÂ am trying to construct a multilevel event history model. Specifically, the
> model is specified as a cross-level main effects model regarding xi as the
> IV at the first level, and z as the IV at the higher level.
> Does anyone know how to write the SPSS syntax? Â To the best of my knowledge,
> SPSS could only handle linear multilevel analysis.
> I have also checked the syntax of STATA. I found that only the clustering
> variable (i.e., groups, clusters), rather than the IVs at the higher
> level could be incorporated to the model by "shared" command.
> Thanks,
> ZHANG Lun

=====================
To manage your subscription to SPSSX-L, send a message to
LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD